The word "existent" seems to be quite Russellian, its not
needed to translate Zermelo. Zermelo rather has a word
for the opposite and more for sets than for classes,
i.e. when a set is empty, he uses the word:
"verschwindet"
https://context.reverso.net/translation/german-english/verschwindet
"existent" would then mean "nicht verschwindet". The problem
with translating "verschwindet", it would translate into verb
"disappear", suggesting that sets are non-rigid modal
placeholders. Although the german language is like that,
and WM causes constantly problems because of that, Zermelo
uses "disappear" adjektively, characterising the
result of an operation.
See for yourself:
Zermelo 1908
https://zenodo.org/record/1428264/files/article.pdf
Because Zermelo characterizes the outcome of operations
on sets, which give an again sets, his axiom of choice is
also not some global choice. Russells dealing with classes
has the danger that one arrives at a different axiom of
choice, than that promoted by Zermelo.